Time To Start Getting REALLY Bullish?
Tom McClellan reads the cycles the morning after the Fed's first hike in three years: a seasonal stock bottom due within about a week, then a year-3 presidential-cycle and year-7 decennial rally; junk-bond breadth as the one thorn; margin debt at a record but its seven-year top not due until ~2028; and gold's 20-month lead saying yields and oil keep rising into 2028.
One-line take: "I am not bullish today, but I am looking for the moment probably within the next week to turn bullish and then I'm going to be bullish as all get out." Stocks are in the late-September/early-October seasonal bottoming window ahead of the presidential cycle's year 3 (up every time since the 20th Amendment except 1939) and a bullish year-7 decennial start; the NYSE A-D line is at highs (worst 3-month drawdown after a 3-year A-D high is ~10%). The caveats are junk-bond breadth (washing out now), record margin debt (7-year cycle top ~2028), and a quiet Warsh slowdown of QE5. He is bearish bonds and bullish oil: gold's moves echo in yields and crude ~20 months later, pointing higher into ~August 2028. Recorded ~1 hour after the Sep-16 FOMC hike. The New Harbor Financial segment (52:52 onward) is the hosts' firm's view, not McClellan's.
1. Stocks & names mentioned
McClellan names no individual companies; the rows below are the asset classes he makes explicit calls on, keyed to the hub's existing index/commodity ids (SPY for US stocks, TLT for long Treasuries, HYG for junk bonds, Oil, Gold). New Harbor's TLT and energy-stock views (from 52:52) are not his and are not rows.
| Ticker | Name | Research | View | What he said | At |
|---|
| SPY | SPDR S&P 500 ETF Trust (US stocks) | QT · SA · STK | Positive | "I am not bullish today, but I am looking for the moment probably within the next week to turn bullish and then I'm going to be bullish as all get out": the S&P is in the late-Sep/early-Oct seasonal bottom before presidential-cycle year 3 (only 1931 and 1939 failed) and a bullish year-7 decennial start; the NYSE A-D line is at highs, which historically caps the next 3 months' worst drawdown near 10%. | 10:56 |
| Crude oil | Crude oil (the molecule) | — | Positive | Gold leads crude by ~19.8 months and "oil prices according to gold still have a lot further to go" — the uptrend is due to last until ~2028; commercials are only lightly net short above $100 ("they don't want to lock in these prices"), another sign of higher prices. "Higher for longer is the bet," sadly. | 32:48 |
| HYG | iShares iBoxx $ High Yield Corporate Bond ETF (junk bonds) | QT · SA · STK | Neutral | The "one thorn": the high-yield bond A-D line has diverged bearishly from prices all 2026 — junk bonds are "horrible investments" that show liquidity pain first — but its McClellan Oscillator is "way the heck down there," wringing out the worst; he wants junk to start doing better in Oct/Nov as confirmation. (HYG is the proxy; he names no fund.) | 15:37 |
| Gold | Gold (as a leading indicator) | — | Neutral | No call on gold itself — he uses it as a clock: its moves echo in 30-year Treasury yields ~20.5 months later and in crude ~19.8 months later; gold's January-2026 top points to a yield/oil peak around August 2028. Open question: does central-bank (China) buying "diminish the message"? | 29:06 |
| TLT | iShares 20+ Year Treasury Bond ETF (long bonds) | QT · SA · STK · FA | Negative | "I'm bearish on bonds right now for a lot of reasons": every QE round (QE1–QE5) has seen bond prices tank, and gold's surge 20.5 months earlier maps to a steep rise in yields from late 2026 into ~2028 — a 30-year above 6% is "reasonable." Only a shift to QT would mitigate it. (TLT is the proxy; he speaks of long Treasuries.) | 27:05 |
2. Talking points
01:36 A strong market: a very upward version of sideways
- Years 1–2 of a presidential term are "supposed to be sideways"; this is "a very, very upward version of sideways."
- Why: low taxation leaves more money in the economy — "we're not eating our seed corn." Problems for the government, "great for the stock market."
02:29 The "Rapunzel chart" — building the presidential cycle
- Chop the market into four-year chunks, reset each to the same starting value, average them. He starts each term on November 1 (election time).
- Years 1–2 sideways, year 3 great, year 4 up but iffier. Year 3 has failed only in 1931 (Depression) and 1939 (the Wehrmacht in Poland).
- Correlation is in the "dance steps," not the slope; exogenous shocks (the Iran war, Hormuz) bend the slope, not the steps.
04:59 The bottom comes before the midterm, not after it
- The market tends to bottom in late September/early October, once investors think they know the election outcome; the outcome itself matters "way far less than the certainty."
- The S&P is running about a week ahead of schedule, in a multiple-bottom structure; the last low should come "somewhere between now and about a week from now."
- What he wants to see at the low: a high put/call day, a high VIX, breadth (and breadth-momentum) divergences, a higher low in the McClellan Oscillator.
08:46 Does this year's strength steal from next year?
- Half of terms beat the average and half lag; with only a handful of cycles since the 20th Amendment, "if you try to squeeze too many nuanced insights out of it, you start running into trouble." Generally "strength tends to continue."
09:47 The decennial pattern and the year-7 effect
- Averaging the Dow in 10-year chunks: a bottom in early October, then year 7, bullish except 1987 — problems in 7-years tend to come later in the year; the bullish start begins in October.
- "That's a screwy idea that it should matter… I gathered the data… and I refuted the hypothesis." Part of it double-counts the four-year cycle (6-years are often term year 2).
12:43 The A-D line: no divergence, and the 3-year-high study
- The NYSE advance-decline line made tops with prices — enough liquidity that "even the lowly stocks" get some.
- After a new 3-year high in the A-D line, the worst drawdown over the next 3 months is typically ~10% (average ~4%). Exceptions: COVID and the abrupt ends of QE1 and QE2.
15:09 The thorn: junk-bond breadth
- His daily A-D line for corporate high-yield bonds has diverged bearishly from prices all 2026; junk "trade much more like stocks than like T-bonds" and draw from the same liquidity pool.
- Its McClellan Oscillator is deeply oversold — the lowest oscillator reading usually precedes the final price low, so "the final price low is probably within a week or so."
- Junk is more sensitive to stock-market liquidity than to rates; watch for junk to improve in October/November as confirmation "there's gobs of money."
20:24 Margin debt: record high, but the seven-year clock says ~2028
- FINRA margin balances (since 1997) are parabolic; normalized by GDP still the highest on record. Peaks tend to come before stock-market tops.
- Peaks recur roughly every seven years (also in quarterly St. Louis Fed FRED data back to 1945); counting from August 2021 points to ~2028 — "five is not seven," so there is "a little bit more room."
24:25 The fly in the ointment: QE5 slowing quietly under Warsh
- By his count we're in QE5; every QE has been bullish for stocks. Since Warsh took over, the slope of Treasury + MBS holdings has flattened without any announcement, and flattened more in the latest data.
- Full-blown quantitative tightening would be the risk — "you got to not just watch the press conference, you got to look at the data."
26:15 QE is bearish for bonds — he is bearish bonds
- Every QE round (2009, 2011, 2012–13, post-COVID, now QE5) saw the bond market tank; bonds went sideways when QE stopped. Counterintuitive, but "at some point, you got to stop arguing with how it should work."
28:04 Oil and rates joined at the hip; gold leads yields by ~20.5 months
- The 10-year has made a higher high while crude hasn't yet exceeded March — people take this spike more seriously than the first.
- Shift gold forward 20.5 months and its dance steps line up with the 30-year yield: a sideways period, then a steep advance in yields from late this year. Direction and timing, not magnitude.
- A 30-year above 6% and a 10-year near 5.5% "sound reasonable"; the pain lands on mortgages/housing (he bought his first house at 13%).
32:48 Gold leads oil by ~19.8 months — higher oil into 2028
- Back to 2014, gold's dance steps reappear in crude ~19.8 months later; today's oil up-move is "right on schedule," with "a lot further to go." Shocks (COVID, Ukraine, Iran) bend the curve temporarily.
- Gold topped January 2026 → the yield and oil peak is due around August 2028. Central-bank buying behind gold's run is the question that "keeps me up at night."
36:54 The COT read: commercials won't lock in $100 oil
- Commercial crude traders (mostly producers hedging) have been net short continuously since 2009; a very high net short marks tops, a low one bottoms.
- They hedged heavily into the first Iran-war spike, then backed off; with oil back above $100 they are "a little bit more timid" — "these experts know something." No price target: "longer to run" is not the same as "more to run."
40:55 The McClellan Market Report and the Oscillator's origins
- Free weekly Chart In Focus; paid newsletter and daily edition. Sherman McClellan (92) still works on it daily; he and Tom's mother built the Oscillator in 1969 by hand on ledgers.
45:58 Why "bullish as all get out"
- The counterarguments fall away: junk breadth can resolve, weak seasonality ends in ~a week and a half, breadth is strong, taxation is low, the Fed "not getting too stupid yet," and year 3 "always works."
- Deficit spending is bullish; a balanced budget would be the big bearish factor — "spending on the credit card makes for a great party."
- 2028 is the convergence date: the seven-year cycle for stocks and margin debt, and the oil/yield top — bad luck for whoever takes office in 2029.
52:52 Host segment: New Harbor Financial (not McClellan)
- Mike Preston agrees the midterm outcome matters less than certainty and that margin debt is a late-cycle signpost (possibly a blow-off first); New Harbor holds ~50% equities and oil-service stocks.
- They disagree on bonds: tactically bullish high-quality US bonds (7–12% of the model; TLT could run toward 120 in a stock sell-off), but not a 10-year hold. John Llodra: averages hide the lost decades (60/40 worst 10-year real return −32%).
- Llodra on the Sep-16 FOMC: a unanimous 12–0 quarter-point hike, the first since July 2023; FedWatch shows zero odds of cuts and ~88% odds of more hikes by December; 10-year touched 5.016%.
3. In plain English
SPY — US stocks (S&P 500) Positive
McClellan is a chart and cycle reader. He averages the stock market over every four-year presidential term and finds the same shape: two flat-ish years, then a strong third year. We are about to enter year 3, and that year has been positive every time for nearly a century except 1939, when World War II began. A ten-year version of the same exercise points the same way for 2027.
The rally usually starts from a low in late September or early October, about a month before the midterm elections, because investors stop worrying once they think they know the result. He is not buying yet: he wants signs of a final washout (a spike in fear gauges, a higher low in his breadth oscillator) within about a week. Breadth — how many stocks are rising versus falling — is already strong, and history says that caps the likely damage in the next three months at about 10%.
Oil — Crude oil Positive
He has found that crude oil tends to repeat gold's ups and downs about 20 months later. Gold surged into January 2026, so on that clock oil should keep trending higher until around mid-2028, with pullbacks along the way. He will not name a price target.
A second check: the "commercials" in the futures market — mostly oil producers who sell future production ahead to lock in a price. When they hedge heavily, prices tend to top; right now, with oil over $100, they are reluctant to lock in, which suggests the people closest to the oil expect higher prices.
HYG — Junk bonds Neutral
Junk bonds are loans to weaker companies that pay a high interest rate because they are risky. McClellan calls them "horrible investments" that only do well when money is plentiful, which makes them an early warning of drying-up liquidity. All year, fewer junk bonds have been rising even as prices held up — a warning sign he calls the one thorn in his bullish case.
But his momentum gauge on that group is now extremely oversold, which usually happens just before the final low. If junk bonds start improving in October and November, he takes it as confirmation that there is plenty of money around and the stock rally is on.
Gold — Gold as a clock Neutral
McClellan makes no call on gold's own price here. He uses it as a leading indicator: whatever gold does tends to show up in long-term interest rates about 20.5 months later and in oil about 20 months later. Because gold peaked in January 2026, he expects rates and oil to peak around August 2028.
His one worry about the signal: this gold rally was driven by central banks (especially China) rather than ordinary investors, which may weaken what it predicts.
TLT — Long-term Treasury bonds Negative
When interest rates rise, the price of existing long-term bonds falls. McClellan expects long-term rates to keep rising: gold's big run 20 months ago maps onto a steep climb in the 30-year yield starting late this year, and he finds a 30-year above 6% reasonable. That hurts long bonds and anyone taking out a mortgage.
Counterintuitively, he also notes that every round of Fed bond-buying ("quantitative easing") has coincided with falling bond prices. Only a switch to the Fed shrinking its holdings would soften his view. (New Harbor, the host's advisory firm, disagrees and is tactically bullish on bonds — that is their view, not his.)
Built from the public YouTube episode (auto-transcript saved in the transcript; fillers removed) — wording is McClellan's and the hosts' own. For personal study — not investment advice. © Thoughtful Money (Adam Taggart) for source material.